Chapter One
Social-cognitive Development
Introduction
Investigating how infants and young children begin to participate in and understand the social world has become a major research paradigm within developmental psychology. Social-cognitive development or developmental social cognition has emerged as an important sub-discipline for a number of reasons. First, there is the intrinsic value of a field which concerns itself with understanding and seeking to explain how children become competent social beings; second, social-cognitive development research constitutes an important source of evidence for complementary areas of developmental psychology (e.g. developmental psycholinguistics, moral development) and third, such research seeks to provide 'baseline' developmental data on social-cognitive abilities for a range of specific problems within applied and clinical developmental fields (e.g. autism and learning impairment).
As adult social cognition research gained prominence (Eiser & Stroebe, 1972) calls for a developmental perspective increased (Forgas, 1981), particularly given the attractiveness of an approach which might fulfil demands for developmental psychology to be more ecologically valid and relevant (Bronfenbrenner, 1979; Neisser, 1976). The promise is that if we can only uncover how young children begin to take part in and cognise the social world, we will not only gain insights into developmental processes, but also into the very structure, function, and processing form of social cognition itself.
The term social cognition began to appear in social psychology and developmental psychology textbooks in the late 1970s and early 1980s in the U.S.A.(Shantz, 1975) and Europe (Forgas, 1981), and over the last 10-15 years boundary conditions for various nuances of the phrase have become established along one or other dichotomies, for example, European/American; rationaVaffective; cognitive/social; Piagetian/ Vygotskian and so on. Thus a cognitive view of social cognition:
' ... an approach that stresses understanding of cognitive processes as a key to understanding complex, purposive, social behaviour' (Isen & Hastorf, 1982, p.2),
can be contrasted with a social outlook:
' ... not merely the information processing analysis of social domains, but [as] a field genuinely devoted to the study of everyday knowledge and understanding ... thejoint product ofsocio-cultural forces and individual cognitive processes' (Forgas, 1981, p.259).
And as an example of a European/American dichotomy, Durkin (1986) notes the European approach as being concerned with social social cognition:
' ... cognition is seen as an achievement promoted and regulated via social interaction'
while the American view emphasises individual social cognition:
'social cognition ... thoughts and reasoning about people, their mental and physical attributes, their personalities, behaviour and viewpoints' (pp.206-207).
We can also note definitions that attempt to negotiate a position somewhere between individuaVsocial extremes, by articulating what is to be taken as social and what as individual, for example:
' ... All the ways in which the child exchanges, receives and processes information from others. These ways include some general cognitive processes, such as attention and memory as well as some that are strictly social, such as communication and perspective taking' (Damon, 1981, p.82).
During the 1970s and early 1980s a number of books and review articles began to mark out a series of research issues which were to form the basis for much of current developmental social-cognitive research (Garvey, 1977; Higgins, Ruble, & Hartup, 1983; Mugny & Doise, 1978; Perret-Clermont, 1980; Schaffer, 1977; Shantz, 1975; Shatz & Gelman, 1973; Turiel, 1982). The emergence of developmental social cognition as a distinct field of inquiry can be traced to a range of interests, some theoretically inspired, such as the cross-fertilisation of models from cognitive psychology; some methodological, for instance driven by the concerns of educationalists and clinicians, and others simply pragmatic (e.g. moral development researchers seeking to incorporate a somewhat marginalised area within a broader theme).
The enthusiasm that this research engendered was related to the growing dissatisfaction for some researchers with Piagetian logico-mathematical accounts of development. Although theoretically sophisticated this approach tended in practice to emphasise what the child couldn't do rather than on her abilities at any given age (Donaldson, 1978; McGarrigle, Grieve, & Hughes, 1978). Another important reason was that the social-cognitive perspective appeared to offer the benefits accruing to the rigour and formalism of the dominant view in psychology, yet within a framework which could and would address developmental social-interactional processes. In other words, the emergence of developmental social cognition as a distinct field of enquiry is interdependent with the widespread utilisation of the information processing metaphor. This is evidenced in the observation that many of the concerns of the field (as well as the underlying constructs) could be located in disparate areas of the behavioural sciences before definitions of social cognition and developmental social cognition began to appear (Cartwright & Zander, 1968; Forgas, 1981; Shantz, 1975; Stotland & Canon, 1972). The application of the cognitive approach to the study of a variety of social-developmental or at least socially oriented concerns, has proceeded rapidly and with some enthusiasm by researchers in child language, cognitive and moral development, educational psychology and learning impairment.
The argument I wish to develop is that the current theoretical frameworks have not fulfilled their earlier promise and by their nature (that is given their pre-theoretical constructs and implicit assumptions), are unlikely to provide adequate concepts and 'tools' necessary for our research purposes. The principal aim of what is to follow then, is to present an alternative framework for studying young children's social cognitive skills. Towards this end, there are three subsidiary objectives. First, developmental social cognition research is critically reviewed (in Chapters 2 and 3), with particular reference to the principal theoretical positions and against a background of considerations and concerns which permit an evaluation of the current 'state of the field'. The specific misgivings outlined (initially in the remainder of this chapter) also underpin a critique of approaches which focus on the young child's social cognitive skills, or attempt to integrate social cognition research with developmental psycho linguistics (reviewed in Chapter 4).
A summary of this literature then serves as the springboard for the second subsidiary objective, outlining and developing an alternative conceptual framework for investigating and understanding young children's active engagement and participation with the social world (in Chapter 5). This approach takes up a number of ideas from within and beyond psychology, notably from conversational analysis, micro-sociology, pragmatics and from the ecological perspective, the notion of affordances. The concern is with understanding what might be required if we wish to formulate richer conceptual models of social-cognitive development.
The third objective is to consider lines of evidence which lend support to the proposed framework (Chapters 6 through 8). The focus is upon one way in which the framework can aid in our understanding one contributory aspect of the child's developing social-cognitive skills-the role of overhearing and its relation to the acquisition of conversation-participation abilities. Finally, the concluding chapter draws together the themes outlined, highlighting various implications of the framework advocated. The overarching goal is to provide a more adequate conceptual basis for developing appropriate models and methods to aid our investigations of social-cognitive development and particularly children's conversationalparticipation skills.
Social-cognitive Development: Problems and Prospects
I would like to distinguish between areas of research activity and primary theoretical orientations. Clearly there are a variety of subjects, applied issues and distinct domains of enquiry which now come under the rubric of developmental social cognition research and the range of topics is fairly broad. Consider one alignment in developmental social cognition which originates in ethologically inspired social interaction research (Bowlby, 1959; Hinde, 1979; Schaffer, 1984). This research holds to the information-processing perspective as one of its (normally implicit) central formulating constructs, combined with ideas from evolutionary biology and ethology. Historically this paradigm made a significant contribution to developmental psychology by showing that the infant was part of a symbiotic network (the 'mutuality' model, Schaffer, 1980) and should not be considered in isolation (Schaffer & Emerson, 1964; Stern, 1974). As long as the orientation here remained 'dyadic', that is, focused upon how the infant affects the parent and the parent the child, the approach was considerably successful. However, as attempts were made to extend the paradigm outwards to the study of 'triadic' and 'polyadic' interactions (the child interacting with three or more people-Liddell & Collis, 1988), the conceptual constraints deriving from an overtly 'meta-descriptive' evolutionary approach became more serious, particularly with reference to the relationship between theory and data (i.e. the proliferation and complexities of the latter which themselves helped clarify the theoretical limitations of the approach).
Another area of developmental social cognition research emerged from what was previously defined as moral development, where now themes, methods and specific topics are often subsumed under the category social-cognitive development. Examples of the 'transformation' of moral into social-cognitive development can be seen in Eisenberg's outline of prosocial reasoning (Eisenberg, 1986) and Hoffman's conception of the development of empathy (Hoffman, 1982). Whether this or that approach derives from innatist conceptions (Turiel, 1981) or maturational-Piagetian views (Kohlberg, 1966) their position or status as developmental social-cognition research is closely tied to the concentration upon performance and measurement criteria (Lefebvre-Pinard, 1982), that is, indices of developing social cognitions. What is interesting about much of this work is that by adopting a social-cognitive perspective the terms of reference appear to have moved away from precedents which were more directly related to pre-theoretical concerns of 'ethics', onto formulations which place the emphasis upon the child's developing understanding or cognitions of social behaviour.
A further important strand arose from the concerns of practitioners in educational and clinical settings. Although one result ofthe Piagetian influence upon educational policy was a move away from the rigours of disciplined rote-learning to more child-centred educational practice, arguably the Piagetian framework is directly applicable only to the more formal areas of the curriculum (science and mathematics). Those concerned with topics somewhat outside of cognitive-developmental domains derived only limited benefit from this approach. In contrast, part of the appeal of the social cognition approach is that here is a representational/constructivist framework with a generally applicable metaphor suitable for a range of educational concerns (e.g. from sociometric indices of peer relations in the classroom through to discourse-based approaches to reading development). It also remains grounded in the requisite formalisms of educational science policy.
Similarly, clinicians and practitioners in learning impairment, mental handicap and remedial language contexts envisage substantial advantages, particularly given that the social-cognitive view provides a terminology of attention deficits, information processing constraints and related terms as a consolidating language, a shared focus for research and homogeneous methodological constructs applicable to a wide variety of child subject groups. The learning-impaired and mental-handicap developmental literature looks towards social cognitive developmental research for additional reasons. First, there is the pertinent need for contributory comparative (baseline) data to assist intervention or development programs. Second, there is the necessity to theoretically understand the relationship between models of development and models of impairment, particularly given that there are very few models of social-cognitive impairment as such (Shute & Paton, 1990). Third, there is an underlying problem with the uncritical adoption of models from 'normal' development paradigms into the study of impairment (e.g. the sometimes unwarranted assumption that by examining impairment and how it changes one necessarily gains insights into how 'normal' processing works). The social-cognitive perspective holds the promise of providing a unifying construct or at least the requisite point of origin for divergent approaches, each sensitive to the concerns of specialist target groups.
Given the pre-eminence of cognitive science in psychology, the benchmark or point of orientation for the majority of current formulations of developmental social cognition research is the information processing framework (in its symbol manipulating representational sense). Arguably while there might be debates between nuances of various positions (e.g. Piagetian vs. Vygotskian) the 'discourse-frame' remains firmly grounded upon the computational metaphor. Consider the way in which the more dominant meaning of the term social cognition has become equated with the application of the information processing metaphor (to the study of social phenomena and processes), noted by Forgas (1981). In his emphasis for a social cognition outlook he argues:
'Cognition, when taken as a domain concerned with all processes of knowing, is intrinsically, inevitably and profoundly social. Our knowledge is socially structured and transmitted from the first day of our life, is coloured by the values, motivations and norms of our social environment in adulthood, and ideas, knowledge and representations are created and recreated at the social as well as the individual level' (p.2).
Beyond the recognition that the use of the word 'transmitted' itself is predicated on a 'computational' discourse, there are a number of points which together might suggest that adherence to the information processing metaphor has led to theoretical impoverishment, that is as far as understanding developmental social cognition as a skill or ability is concerned.
The first reason derives from a commonly found distinction in psychology: some fields of enquiry are to be properly addressed as individualistic (attention and memory) while others (communication) are intrinsically social, and on few occasions is there any attempt to remind ourselves that this differentiation is based upon theoretical assumptions, which are themselves controversial and need to be substantiated. Given that memory and learning how to remember may itself be established in social-discursive practices (Edwards & Middleton, 1988), as well as the fairly extensive research highlighting the social nature of categorisation (Tajfel & Forgas, 1981; Zajonc, 1980), while there may not yet be sufficient grounds for abandoning such a distinction, there is an increasing requirement for theoretical justification.
The second is the inherent danger of arguing that all cognition is intrinsically social while at the same time maintaining that there remains an important and uniquely essential 'individual level'. It has been argued elsewhere that part of the reason why cognitive psychology is largely viewed as non-social is because language is not considered as a fundamentally social phenomena (Farr, 1981); that is, the emphasis has been on language as a formal symbol system. This of course is not surprising given that the birth of psycholinguistics can be traced to the publication of Chomsky's (1957) monograph on competence grammars. Some theoreticians in cognitive science have made serious attempts to highlight the limitations of abstract and formally complex models of language (e.g. Givon, 1984), pointing out the inherent dangers of strictly formal 'rationalistic' approaches to language understanding. However, a more serious problem revolves around a presupposed relationship between language and identity.
This issue arises with reference to the philosophical and conceptual basis for the scientific status of concepts, categories and those various hypothesis regarding the nature of language and thought (Chomsky, 1957; Rosch, 1975; Saussure, 1966; Whorf, 1956). If one holds to a cognitive/computational conception (Chomsky, 1988; Fodor, 1983) of language and cognition, then beyond the various debates on precisely how to model formal rules acting upon representations of information processing activities (Clark, 1987; Karmiloff-Smith, 1979; Pinker, 1984; Rumelhart & McClelland, 1986) the underlying agenda locates thought as being first and foremost 'inside the individual's head' and an inherent essential property of that individual. The point here is that whatever particular conception of cognition or language understanding that follows has placed the issue of 'identity' (of the cogito or logos) as the central question. All consequential attempts to adequately address 'interaction' dimensions are already compromised by the metaphysics of this position.
In contrast to such post-Kantian formulations (of cognition) other threads of continental philosophy (particularly Hegel, Merleu-Ponty, and Derrida) have sought to demonstrate that a number of philosophical arguments concerning identity (and representation), based upon a priori categorisation, contain a number of serious inherent contradictions. These serve to make them unworkable or at least severely constrained by outdated conceptions of 'grand truths' and the centrality of identity (Descombes, 1980). This is not the place to enter into the ramifications of recent philosophical developments for psychology-problematic though they may be-see Feldman and Bruner (1987) and Stitch (1983), only that with regard to language there are now serious grounds for arguing that not only are many models within cognitive psycholinguistics unhelpful but that there are alternative emerging theoretical frameworks which may be better placed to provide the basis for a social social cognition (e.g. Sinha, 1988).
A third reason for calling into question the appropriateness of the current theoretical orientation is the emphasis on one of the essential building blocks of the current framework supporting social-cognitive theories, the developmental dimension itself (Durkin, 198...